Automatic identification of systematic repeating defects in semiconductor production
    4.
    发明授权
    Automatic identification of systematic repeating defects in semiconductor production 有权
    自动识别半导体生产中的系统重复缺陷

    公开(公告)号:US08312395B2

    公开(公告)日:2012-11-13

    申请号:US13007556

    申请日:2011-01-14

    IPC分类号: G06F17/50

    摘要: A method includes capturing an image of the pattern using one or more scans across a surface of the partially completed wafer. The method includes processing information associated with the captured image of the pattern in a first format (e.g., pixel domain) into a second format, e.g., transform domain. The method includes determining defect information associated with the image of the pattern in the second format and processing the defect information (e.g., wafer identification, product identification, layer information, x-y die scanned) to identify at least one defect associated with a spatial location of a repeating pattern on the partially completed wafer provided by a reticle. The method includes identifying the reticle associated with the defect and a stepper associated with the reticle having the defect and ceasing operation of the stepper. The damaged reticle is replaced, and the process resumes using a replaced reticle.

    摘要翻译: 一种方法包括使用在部分完成的晶片的表面上的一次或多次扫描来捕获图案的图像。 该方法包括以与第一格式(例如,像素域)中的图案的捕获图像相关联的信息处理成第二格式,例如变换域。 该方法包括确定与第二格式的图案的图像相关联的缺陷信息,并处理缺陷信息(例如,晶片识别,产品识别,层信息,xy模扫描),以识别与空间位置相关联的至少一个缺陷 在由光罩提供的部分完成的晶片上的重复图案。 所述方法包括识别与所述缺陷相关联的掩模版和与所述掩模版相关联的具有所述缺陷并且所述步进器停止操作的步进器。 损坏的标线被更换,并且使用更换的掩模版恢复工艺。

    Method for measuring confidence of ball grid array model in surface mounted devices
    8.
    发明申请
    Method for measuring confidence of ball grid array model in surface mounted devices 有权
    用于测量表面安装装置中球栅阵列模型的置信度的方法

    公开(公告)号:US20040213451A1

    公开(公告)日:2004-10-28

    申请号:US10420286

    申请日:2003-04-22

    发明人: Tong Fang Ming Fang

    IPC分类号: G06K009/00

    摘要: A method for obtaining confidence measure of a ball grid array (BGA) model having a plurality of balls in semiconductor surface mounted devices is provided. The method comprises the steps of extracting BGA images from a real surface mounted device, generating a BGA ball model and a BGA body model, generating a first confidence measure of the BGA ball model wherein the first confidence measure includes a first standard deviation of the BGA ball model and a first local image contrast of each BGA ball, and generating a second confidence measure of the BGA body model wherein the second confidence measure includes a second standard deviation of the BGA body model and a second local image contrast of the BGA body.

    摘要翻译: 提供了一种用于获得半导体表面安装器件中具有多个球的球栅阵列(BGA)模型的置信度测量的方法。 该方法包括以下步骤:从真实表面安装的装置提取BGA图像,产生BGA球模型和BGA体模型,产生BGA球模型的第一置信度,其中第一置信度测量包括BGA的第一标准偏差 球模型和每个BGA球的第一局部图像对比度,并且生成BGA体模型的第二置信度测量,其中第二置信度测量包括BGA体模型的第二标准偏差和BGA体的第二局部图像对比度。

    Method of and apparatus for inspecting transparent object for defect
    9.
    发明授权
    Method of and apparatus for inspecting transparent object for defect 失效
    检查透明物体缺陷的方法和装置

    公开(公告)号:US5216481A

    公开(公告)日:1993-06-01

    申请号:US808620

    申请日:1991-12-17

    申请人: Nobuhiro Minato

    发明人: Nobuhiro Minato

    摘要: A method of and an apparatus for inspecting a transparent object for a defect wherein both of a light blocking defect and a refracting defect can be discriminated not only for presence or absence thereof but also for a shape and a kind thereof and besides discrimination of a small defect can be performed with a high degree of accuracy. According to the method, a pitch of strips of a reference striped pattern is detected in prior. Then, an object for inspection is placed at an inspecting position, and light having the reference pattern is projected upon the object and transmission light is photographed by an image sensor. A threshold value is set from an average value between two picture element data spaced from each other by one half the detected pitch, and the picture element data are successively compared with the threshold value to determine the bright or the dark thereof. A defect of the object is discriminated from numbers of picture elements determined as the bright and the dark.